arXiv subjects

Alfonsas Juršėnas

Publications and source records attributed to Alfonsas Juršėnas.

4 recordsLinked to original sources

arXiv2023

Optical Bessel-like Beams with Engineered Axial Phase and Intensity Distribution

Bessel beams are known for their property of maintaining propagation-invariant transverse intensity distribution. The true Bessel beams has unlimited energy, however the experimental realization of limited energy Bessel-like beams provides almost diffraction-free beams over a certain distance. These beams are applicable to microfabrication, particle manipulation etc. where long and narrow intense beam is needed. A type of laser beams that maintains narrow intensity peak for long distances are often called optical needles. In this paper we propose a construction of an optical needle using a superposition of vector Bessel beams and for flexible engineering of axial intensity profile. We investigate both theoretically and experimentally generation of various optical beams of different axial profiles, lengths, beam widths etc. We use a spatial light modulator as a toy model of an optical element to shape the spatial spectrum of the beam. Moreover, we demonstrate distortions introduced into the optical needle by a planar interface between two dielectric media: spherical aberration and astigmatism due to the nonzero inclination angle and discuss ways to correct them.

physics.optics↗

arXiv2023

Controllable Spatial Array of Bessel-like Beams with Independent Axial Intensity Distributions for Laser Microprocessing

Bessel beams generated via axicons are widely used for various applications like optical tweezers or laser microfabrication of transparent materials. The specific intensity profile having high aspect ratio of beam width and length in turn generates high aspect ratio void that resembles a needle. In contrast to commonly generated Bessel beam that has a fixed axial intensity distribution. We present a novel method to engineer an optical needle that can have an arbitrary axial intensity distribution via superposition of different cone angle Bessel beams. We analytically describe spatial spectra of an optical needle having arbitrary axial intensity distribution. We also demonstrate a superposition of independent optical needles and analyze the physical limitations to observe well separated optical needles as they are influence by mutual interference of the individual beams. In order to verify our theoretical and numerical results we generate controllable spatial arrays of individual beams with various numbers and spatial separations by altering a spectrum of incoming laser beam via spatial light modulator. Lastly, we numerically examine distortions caused by propagation through planar air-dielectric interface and show compensation method by appropriately modifying spectral masks.

physics.optics↗

arXiv2022

Automatic travel pattern extraction from visa page stamps using CNN models

Manual travel pattern inference from visa page stamps is a time consuming activity and constitutes an important bottleneck in the efficiency of traveler inspection at border crossings. Despite efforts to digitize and record the border crossing information into databases, travel pattern inference from stamps will remain a problem until every country in the world is incorporated into such a unified system. This could take decades. We propose an automated document analysis system that processes scanned visa pages and automatically extracts the travel pattern from detected stamps. The system processes the page via the following pipeline: stamp detection in the visa page; general stamp country and entry/exit recognition; Schengen area stamp country and entry/exit recognition; Schengen area stamp date extraction. For each stage of the proposed pipeline we construct neural network models and train then on a mixture of real and synthetic data. We integrated Schengen area stamp detection and date, country, entry/exit recognition models together with a graphical user interface into a prototype of an automatic travel pattern extraction tool. We find that by combining simple neural network models into our proposed pipeline a useful tool can be created which can speed up the travel pattern extraction significantly.

cs.CV↗

arXiv2020

Training Deep Spiking Neural Networks

Computation using brain-inspired spiking neural networks (SNNs) with neuromorphic hardware may offer orders of magnitude higher energy efficiency compared to the current analog neural networks (ANNs). Unfortunately, training SNNs with the same number of layers as state of the art ANNs remains a challenge. To our knowledge the only method which is successful in this regard is supervised training of ANN and then converting it to SNN. In this work we directly train deep SNNs using backpropagation with surrogate gradient and find that due to implicitly recurrent nature of feed forward SNN's the exploding or vanishing gradient problem severely hinders their training. We show that this problem can be solved by tuning the surrogate gradient function. We also propose using batch normalization from ANN literature on input currents of SNN neurons. Using these improvements we show that is is possible to train SNN with ResNet50 architecture on CIFAR100 and Imagenette object recognition datasets. The trained SNN falls behind in accuracy compared to analogous ANN but requires several orders of magnitude less inference time steps (as low as 10) to reach good accuracy compared to SNNs obtained by conversion from ANN which require on the order of 1000 time steps.

cs.NE↗